Answers are grounded in this paper and cite their sources.
Weakly ferromagnetic alloys are being used by several groups in the study of superconducting/ferromagnetic hybrid systems. Because spin-flip and spin-orbit scattering in such alloys disrupt the penetration of pair correlations into the ferromagnetic material, it is desirable to have a direct measurement of the spin memory length in such alloys. We have measured the spin memory length at 4.2 K in sputtered Pd0.88Ni0.12 and Pd0.987Fe0.013 alloys using methods based on current-perpendicular-to-plane giant magnetoresistance. The alloys are incorporated into hybrid spin valves of various types, and the spin memory length is determined by fits of the Valet-Fert spin-transport equations to data of magnetoresistance vs alloy thickness. For the case of PdNi alloy, the resulting values of the spin memory length are lsfPdNi=2.8±0.5 nm and 5.4±0.6 nm, depending on whether or not the PdNi is exchange biased by an adjacent Permalloy layer. For PdFe, the spin memory length is somewhat longer, lsfPdFe=9.6±2 nm, consistent with earlier measurements indicating lower spin-orbit scattering in that material. Unfortunately, even the longer spin memory length in PdFe may not be long enough to facilitate observation of spin-triplet superconducting correlations predicted to occur in superconducting/ferromagnetic hybrid systems in the presence of magnetic inhomogeneity.
No takes yet. Share an insight, caveat, or question.
Arham et al. (2009) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: